Architecture of Pd–Au Bimetallic Nanoparticles in Sodium Bis(2

Architecture of Pd–Au Bimetallic Nanoparticles in Sodium Bis(2

ARTICLE
Architecture of Pd–Au Bimetallic
Nanoparticles in Sodium Bis(2ethylhexyl)sulfosuccinate Reverse
Micelles As Investigated by X-ray
Absorption Spectroscopy
Ching-Hsiang Chen,†,§ Loka Subramanyam Sarma,† Jium-Ming Chen,† Shou-Chu Shih,† Guo-Rung Wang,†
Din-Goa Liu,‡ Mau-Tsu Tang,‡ Jyh-Fu Lee,‡ and Bing-Joe Hwang†,‡,*
†
Nanoelectrochemistry Laboratory, Department of Chemical Engineering, National Taiwan University of Science and Technology, Taipei 106, Taiwan, ‡National Synchrotron
Radiation Research Center, Hsinchu 300, Taiwan, and §Institute of Atomic and Molecular Sciences, Academia Sinica, Taipei 115, Taiwan
ABSTRACT In this study, we demonstrate the unique application of X-ray absorption spectroscopy (XAS) as
a fundamental characterization tool to help in designing and controlling the architecture of Pd–Au bimetallic
nanoparticles within a water-in-oil microemulsion system of water/sodium bis(2-ethylhexyl)sulfosuccinate (AOT)/
n-heptane. Structural insights obtained from the in situ XAS measurements recorded at each step during the
formation process revealed that Pd–Au bimetallic clusters with various Pd–Au atomic stackings are formed by
properly performing hydrazine reduction and redox transmetalation reactions sequentially within water-in-oil
microemulsions. A structural model is provided to explain reasonably each reaction step and to give detailed
insight into the nucleation and growth mechanism of Pd–Au bimetallic clusters. The combination of in situ XAS
analysis at both the Pd K-edge and the Au LIII-edge and UV–vis absorption spectral features confirms that the
formation of Pd–Au bimetallic clusters follows a (Pdnuclei–Austack)–Pdsurf stacking. This result further implies that
the thickness of Austack and Pdsurf layers may be modulated by varying the dosage of the Au precursor and
hydrazine, respectively. In addition, a bimetallic (Pd–Au)alloy nanocluster with a (Pdnuclei–Austack)–(Pd–Aualloy)surf
stacking was also designed and synthesized in order to check the feasibility of Pdsurf layer modification. The
result reveals that the Pdsurf layer of the stacked (Pdnuclei–Au)stack bimetallic clusters can be successfully modified
to form a (Au–Pd alloy)surf layer by a co-reduction of Pd and Au ions by hydrazine. Further, we demonstrate the
alloying extent or atomic distribution of Pd and Au in Pd–Au bimetallic nanoparticles from the derived XAS
structural parameters. The complete XAS-based methodology, demonstrated here on the Pd–Au bimetallic system,
can easily be extended to design and control the alloying extent or atomic distribution, atomic stacking, and
electronic structure to construct many other types of bimetallic systems for interesting applications.
KEYWORDS: bimetallic nanoparticles · X-ray absorption spectroscopy · redox
transmetalation reaction · reverse micelles · atomic distribution
B
*Address correspondence to
[email protected].
Received for review April 30, 2007
and accepted July 31, 2007.
Published online August 25, 2007.
10.1021/nn700021x CCC: $37.00
© 2007 American Chemical Society
114
imetallic nanoparticles (NPs) in the
form of either alloys or core–shell
nanostructures are of great interest1
from both scientific and technological perspectives because of their attractive physical
and chemical properties, resulting from not
only size effects but also the combination of
different metals.2,3 It was observed that the
activity and selectivity of the metal catalysts
can be drastically influenced by the presence
of a second metal component.4 The “ligand
effect”, which refers to the electronic struc-
VOL. 1 ▪ NO. 2 ▪ CHEN ET AL.
ture modifications resulting from the addition of one metal to a second metal, leads to
the formation of heterometallic bonds involving either charge transfer between the metals or orbital rehybridization of the metals.
These changes can alter the properties of the
mixed-metal system, and various bimetallic
systems formed accordingly have been
reported.5–7 Among the various bimetallic
nanoclusters (NCs), Pd–Au has been widely
explored as a catalytic material for a variety
of reactions, including vinyl acetate
synthesis,8–10 hydrodesulfurization of
thiophene,11 acetylene hydrogenation,12
and oxidation of alcohols to aldehydes.13
Most of the numerous recent reports have focused on the synthesis and catalytic applications of Pd–Au bimetallic NCs.14–19 Although
significant contributions have been devoted
to the synthesis and characterization of
Pd–Au bimetallic clusters, the atomic distribution of the NCs, structural dynamics, and
thermodynamic behavior, which are the key
parameters for catalyst design and architecture, remain less explored.
We have recently reported that the properties of bimetallic catalysts can be significantly influenced by the extent of alloying
between the two constituent elements,
based on X-ray absorption spectroscopy
(XAS) methodology.20 In order to control
properties such as atomic distribution, stacking order, and surface composition, which are
crucial for the architecture of bimetallic NPs
with tailored structure and functions, understanding the formation mechanism remains a
major scientific interest. The commonly available X-ray diffraction (XRD) technique provides information about the long-range orwww.acsnano.org
www.acsnano.org
ter getting detailed knowledge of the reaction
mechanism, we tried to design a (Pd–Au)alloy bimetallic NP with three stacking regions, viz., (Pdnuclei–
Au)stack–(Au–Pd alloy)surf, by controlling two HRRs
and one RTM. It is believed that the detailed knowledge about the formation mechanism gained from
in situ XAS in this study can be further utilized to design and modify the alloying extent, atomic distribution, stacking order, and electronic configuration of
other bimetallic NPs for interesting applications.
ARTICLE
dering and periodicities, preferably of single crystals or
polycrystals, and lacks the means to identify short-range
ordering information; hence, conclusions about the alloy
structure of nanosized particles cannot be drawn simply
from XRD. In fact, it seems that anomalous wide-angle
X-ray scattering is able to give the same information on
the distribution of the two metals inside bimetallic clusters.21 Techniques like transmission electron microscopy
(TEM) and high-resolution (HR) TEM provide only a qualitative understanding of the structure of bimetallic NPs.
For example, the HRTEM method enables one to distinguish the core–shell structure on the basis of observing
different lattice spacing. On the other hand, TEM emphasizes the intensity contrast in understanding the structure
of bimetallic NPs. One method suitable for the study of
the structure of metal clusters is in situ XAS, comprising
both X-ray absorption near-edge spectroscopy (XANES)
and extended X-ray absorption fine structure spectroscopy (EXAFS) regions.22,23 XANES can reveal the oxidation state and d-band occupancy of a specific atom, as
well as the size of metallic clusters,24,25 while EXAFS provides a powerful tool for the analysis of local atomic structure, giving accurate information about the average local
atomic environment. EXAFS is particularly sensitive to interatomic distances and local disorder and has been successfully utilized to resolve subtle structural details about
NPs.26 XAS has been proved as a powerful technique for
the characterization of bimetallic catalysts,27–29 since it is
difficult to obtain structural information on such systems
by conventional material analysis methods at the early
stages.
We have recently made some significant contributions that highlight the applicability of XAS to study nanoparticle formation mechanisms.30–33 We also applied
XAS to study the chemical transformation of bimetallic
NPs to ternary metallic NPs by a cation redox reaction.34
In this study, we further advance the applicability of XAS
to design and construct Pd–Au bimetallic NPs with various Pd–Au atomic stackings in water-in-oil microemulsions by properly employing hydrazine reduction reactions (HRRs) together with redox transmetalation
reactions (RTMs). In a RTM process, the sacrificial oxidation of one metal results in the deposition of another
metal ion to produce bimetallic NPs.35 The thickness of
the gold and palladium layers and their surface composition can be controlled by adjusting the dosage of the gold
precursors and the reducing agent, respectively. We know
that ethylene dehydrogenation can be catalyzed by a Pd
cluster, especially on the surface of Pd(111) and Pd(100)
planes.36,37 It was reported that the addition of Au to Pd
clusters can lead to an enhancement of ethylene dehydrogenation38 due to the sp- and d-band charge redistribution or orbital hybridization, indicating that controlling
the atomic ratio NAu/NPd would further control the electronic perturbation of Au and Pd. We also attempted to
provide details on the atomic distribution and structural
insights into Au–Pd bimetallic clusters, since the catalytic
activity of bimetallic clusters is strongly influenced by
their atomic distribution and structure. Furthermore, af-
RESULTS AND DISCUSSION
In Situ XANES at Various Stages in the Formation of Bimetallic
Pd–Au Clusters with Various Atomic Stackings. XANES can provide information on the electronic transition from the
core level to the vacant d states of the absorbing atom
and is recognized as a powerful tool to characterize in
situ d-electron configuration in catalysts. The Pd K-edge
XANES spectrum is recorded for the microemulsion system containing Pd2⫹ ions as a function of dosage of reducing agent, hydrazine (N2H5OH), in the hydrazine reduction reactions and as a function of NAu/NPd ratio in
the redox transmetalation reactions. The corresponding
Pd K-edge XANES spectra of Pd–Au bimetallic nanoparticles with various stackings are shown in Figure 1.
In the Pd K-edge XANES spectra, the absorption transition in the pre-edge region is a 1 s to 4 d dipoleforbidden transition according to the selection rule, ⌬l ⫽
1 and ⌬j ⫽ 1, where l and j are orbital angular momentum
and total angular moment of the local density of states,
respectively.39 The absorption threshold resonance,
called a white line, appearing between 24360 and 24380
eV, corresponds to the electronic transitions that arise
from the 1 s to unoccupied 4 p states above the Fermi
level. The second (24385 eV) and third (24435 eV) peaks
are attributed to 1 s ¡ dp and 1 s ¡ dsp transitions, respectively. This white line intensity is sensitive to the
change in electron occupancy of the valance orbital and
ligand field environments of an absorber.40 For this reason, it can be used to estimate the density of the unoccupied states and the changes in the oxidation state of the
Pd absorber. The energy shift can also be related to a
change in the effective number of valence electrons at
the absorbing site. As can be seen from Figure 1, the edge
energy of the initial Pd2⫹ solution gradually shifts toward the reference Pd foil, with an increase in the amount
of reducing agent from 0 to 44 ⫻ 10⫺5 mol, indicating
the progressive reduction of Pd2⫹ to Pd0 (Pdnuclei NPs).
For a better understanding, we also provide more detailed Pd K-edge XANES spectra, recorded during the formation of Pd–Au bimetallic NPs, in the Supporting Information, in which a gradual decrease in white line intensity
can also be observed with increasing amount of reducing agent from 0 to 44 ⫻ 10⫺5 mol (Figure 1S). After the
formation of Pd clusters, the Au precursor solution, i.e., the
HAuCl4 emulsion, was introduced, and the hydrazine dosage was increased before the HRRs were performed. The
stability of the state of the system was checked by comVOL. 1 ▪ NO. 2 ▪ 114–125 ▪ 2007
115
ARTICLE
Figure 2. Normalized XANES spectra near the Au LIII-edge at
various steps during the formation of Pd–Au bimetallic clusters with various atomic stacking in reverse micelles, HAuCl4,
and reference Au foil.
Figure 1. Normalized XANES spectra near the Pd K-edge at
various steps during the formation of Pd–Au bimetallic clusters with various atomic stacking in reverse micelles, with
K2PdCl4 and Pd foil as references: (a) Pdnuclei, (b) (Pdnuclei–
Austack-1), (c) (Pdnuclei–Austack-1)–Pdsurf, (d) (Pdnuclei–Austack-2),
(e) (Pdnuclei–Austack-2)–Pdsurf, (f) (Pdnuclei–Austack-3), and (g)
(Pdnuclei–Austack-3)–Pdsurf NPs.
paring two sequential X-ray absorption spectra before
proceeding to the next stage.
After the addition of the Au3⫹ precursor (stoichiometric ratio NAu/NPd ⫽ 0.633) at 5.5 h (Pdnuclei–Austack-1 NPs),
the white line intensity increases slightly due to the oxidation of part of the Pdnuclei NPs. The edge energy is positioned more closely to the K2PdCl4 microemulsion, indicating the formation of Pd–Cl species at this stage, which
will be further confirmed during the discussion of EXAFS
results. After the addition of 79.2 ⫻ 10⫺5 mol of hydrazine
at 6.8 h, the edge position approaches the reference Pd
foil, indicating the reduction of these PdCl42⫺ species
back to Pd0 ((Pdnuclei–Austack-1)–Pdsurf NPs). The edge shift
to K2PdCl4 during the formation of (Pdnuclei–Austack-2)
and (Pdnuclei–Austack-3) NPs by RTMs is consistent with
the oxidation of Pd clusters, whereas the edge shift toward the reference Pd during the formation of (Pdnuclei–
Austack-2)–Pdsurf and (Pdnuclei–Austack-3)–Pdsurf NPs by
HRRs indicates the reduction of Pd2⫹ ions.
The Au LIII-edge XANES spectra of the bimetallic
Au–Pd microemulsion system are shown in Figure 2.
The white line intensity of HAuCl4, which is positioned
around 11923 eV,41 i.e., 4 eV above the edge of the Au
LIII-edge, is sharply decreased and matched well with
the magnitude of white line intensity of the reference
Au foil soon after the HAuCl4 microemulsion is introduced into the Pd microemulsion complex containing
Pd clusters at 6.0 h. This observation indicates that Au3⫹
116
VOL. 1 ▪ NO. 2 ▪ CHEN ET AL.
ions are reduced to Au0 after interacting with Pd clusters. The white line intensity in the Au LIII-edge XANES
spectra shows no change, but the intensity of peak A increases slightly, indicating the possibility of a new interaction between Au and Pd atoms after the addition of
79.2 ⫻ 10⫺5 mol of hydrazine during the formation of
(Pdnuclei–Austack-1)–Pdsurf NPs. When the ratio of NAu/
NPd is increased to 0.862 (Pdnuclei–Austack-2 NPs), the
white line intensity is not changed when compared to
the white line intensity of Pdnuclei–Austack-1 NPs,
whereas the intensity of peak A is lowered, indicating
a discernible difference in the d-bands of the Au core
holes. It would further imply that the electronic configurations of clusters formed at these two stoichiometric
ratios are different. Further increase in the NAu/NPd
ratio to 0.997 does not change the white line intensity,
indicating that all the Au3⫹ ions are reduced to Au0
during the formation of Pdnuclei–Austack-3 NPs.
In Situ EXAFS at Various Stages in the Formation of Bimetallic
Pd–Au Clusters with Various Atomic Stackings. The Fouriertransformed (FT) k3-weighted EXAFS spectra (⌬k ⫽ 4.05–
14.10 Å⫺1) of the microemulsion system at various HRRs
and RTMs are shown in Figure 3. As can be seen from Figure 3, the K2PdCl4 emulsion exhibits a peak between 1.5
and 2.0 Å, corresponding to a Pd–Cl bond. After the addition of 22 ⫻ 10⫺5 mol of N2H5OH at 2.3 h, the magnitude
of Pd–Cl bond is decreased, and a new peak appears between 2.0 and 3.0 Å, attributed to a Pd–Pd bond, indicating that PdCl42⫺ ions are reduced and Pd clusters are
formed directly without any intermediate. The corresponding chemical reaction can be written as shown in eq 1.
N2H5OH + PdCl42-+2H+ f
N2 + H2 + Pd0 + 4HCl + H2O (1)
www.acsnano.org
After the addition of 44 ⫻ 10⫺5 mol of N2H5OH at 4.5 h,
the FT peak corresponding to the Pd–Cl bond disappears
completely, and the magnitude of the Pd–Pd bond increases. This implies that all of the palladium ions are completely reduced to Pd0 and transformed to Pd clusters
(Pdnuclei NPs). After the addition of the Au3⫹ emulsion at
5.5 h (NAu/NPd ⫽ 0.633), the magnitude of the Pd–Pd
bond is decreased and shifted slightly to higher R values
due to the presence of Au. A peak corresponding to a
Pd–Cl bond, which has a magnitude slightly less than that
of the equivalent peak in the EXAFS spectrum of the
PdCl42⫺ precursor, appears between 1.5 and 2.0 Å, indicating that the oxidation of Pd clusters takes place during the redox transmetalation process (Pdnuclei–Austack-1
NPs).
The appearance of the Pd–Cl bond is consistent with
our previous study, in which we showed that the reaction between monometallic Pd clusters with the microemulsion containing Pt4⫹ ions generated the Pd–Cl bond
due to the oxidation of Pd clusters.33 By increasing the total amount of hydrazine to 79.2 ⫻ 10⫺5 mol at 6.8 h
((Pdnuclei–Austack-1)–Pdsurf NPs), the magnitude of the
Pd–Cl bond is decreased, and a new broad peak attributed to a Pd–M bond is formed. The FEFF7 software package was used to fit the Pd–Pd and Pd–Au bonds, as
shown in Table 1. After further addition of the Au precursor (NAu/NPd ⫽ 0.862) at 8.6 h (Pdnuclei–Austack-2 NPs), the
magnitude of the broad peak of the Pd–M bond decreases, whereas the magnitude of the Pd–Cl bond increases. Upon addition of the reducing agent to 96.8 ⫻
www.acsnano.org
ARTICLE
Figure 3. FT-EXAFS spectra obtained at the Pd K-edge at
various stages in the formation of bimetallic Pd–Au NPs with
various atomic stackings and reference Pd foil: (a) Pdnuclei,
(b) (Pdnuclei–Austack-1), (c) (Pdnuclei–Austack-1)–Pdsurf, (d)
(Pdnuclei–Austack-2), (e) (Pdnuclei–Austack-2)–Pdsurf, (f) (Pdnuclei–
Austack-3), and (g) (Pdnuclei–Austack-3)–Pdsurf NPs.
10⫺5 mol at 10.4 h ((Pdnuclei–Austack-2)–Pdsurf NPs), the
magnitude of not only the Pd–Pd bond but also the
Pd–Au bond is increased, and the peak corresponding to
the Pd–Cl is eliminated. After equilibrium is reached at
14.7 h, further Au precursor is added to make equal the
atomic ratio between the Pd and Au atoms (Pdnuclei–
Austack-3 NPs). A similar phenomenon is observed in the
first and second additions of the Au precursors. After addition of reducing agent to 110 ⫻ 10⫺5 mol at 16.5 h ((Pdnuclei–Austack-3)–Pdsurf NPs), the contribution from Pd–Cl
bonding completely disappears.
The FT k3-weighted EXAFS spectra (⌬k ⫽ 3.39 –13.25
⫺1
Å ) obtained at the Au LIII-edge of the bimetallic Pd–Au
microemulsion system are shown in Figure 4. It is found
that the feature of the Fourier transform spectra of the
Au absorber for various atomic ratios NAu/NPd (for the NPs
obtained during all the RTMs, i.e., Pdnuclei–Austack-1,
Pdnuclei–Austack-2, and Pdnuclei–Austack-3) exhibits two
bonds between 1.5 and 3.5 Å in length, the magnitude
of which is increasing with the Au stacking. After fitting
of the k3-weighted EXAFS data by using the FEFF7 software package, the two bonds are evaluated as the Au–Pd
and Au–Au bonds, respectively. As can be seen in Figure 4, the Au–Cl bond is not found during all the
RTMs, and the figure shows only the Au–metal bonds
after the addition of the Au precursors, indicating
that all of the AuCl4⫺ ions are reduced completely
via RTM. The structural parameters (coordination
number N, bond distance R, Debye–Waller factor ␴2,
and inner potential shift ⌬E0) derived from the Au
LIII-edge EXAFS data analysis of Pd–Au bimetallic system are shown in Table 2. The results of Pd K-edge
Figure 4. FT-EXAFS spectra obtained at the Au LIII-edge at
various stages of bimetallic Pd–Au NPs formation in sodium
bis(2-ethylhexyl)sulfosuccinate (AOT) reverse micelles at
various NAu/NPd ratios, as a function of reducing agent dosage, and reference compounds HAuCl4 and Au foil.
VOL. 1 ▪ NO. 2 ▪ 114–125 ▪ 2007
117
ARTICLE
TABLE 1. Structural Parameters Derived from the Pd K-Edge EXAFS Data Analysis at Various Stages in the Formation of
Pd–Au Bimetallic Clusters with Various Atomic Stackings and the Reference Pd Foila
reaction
state
amount of
N2H5OH (mol)
time (h)
atomic ratio
(NAu/NPd)
shell
N
Rj (Å)
␴2
(ⴛ10ⴚ3 Å2)
0
0
0
Pd–Cl
4.00 (0.13)
2.320 (0.003)
4.8 (0.3)
4.5 (0.3)
0.0089
⫺5
2.3
0
Pd–Cl
Pd–Pd
1.86 (0.12)
6.61 (0.08)
2.329 (0.003)
2.743 (0.002)
8.9 (0.1)
7.8 (0.1)
6.9 (0.4)
⫺1.1 (0.1)
0.0021
44 ⫻ 10⫺5
4.5
0
Pd–Pd
8.90 (0.06)
2.742 (0.001)
6.8 (0.3)
⫺1.4 (0.3)
0.0022
5.5
0.633
Pd–Cl
Pd–Pd
Pd–Au
2.84 (0.13)
2.17 (0.08)
0.21 (0.08)
2.314 (0.003)
2.742 (0.002)
2.757 (0.003)
4.8 (0.2)
8.4 (0.1)
0.0 (0.2)
1.2 (0.5)
⫺4.1 (0.3)
13.0 (0.8)
0.0036
Pd–Pd
Pd–Au
6.37 (0.11)
0.87 (0.07)
2.749 (0.004)
2.755 (0.002)
7.5 (0.3)
5.6 (0.3)
⫺3.9 (0.2)
0.1 (0.3)
0.0196
Pd–Cl
Pd–Pd
Pd–Au
2.53 (0.13)
2.59 (0.06)
0.24 (0.09)
2.313 (0.003)
2.742 (0.002)
2.757 (0.002)
5.3 (0.3)
5.9 (0.2)
0.0 (0.2)
0.1 (0.6)
⫺5.1 (0.2)
12.1 (0.5)
0.0033
Pd–Pd
Pd–Au
6.46 (0.07)
1.06 (0.08)
2.749 (0.002)
2.755 (0.004)
6.3 (0.1)
4.3 (0.2)
⫺2.8 (0.3)
0.9 (0.2)
0.0083
Pd–Cl
Pd–Pd
Pd–Au
1.96 (0.14)
1.81 (0.07)
0.21 (0.09)
2.310 (0.003)
2.742 (0.002)
2.757 (0.002)
4.1 (0.3)
4.4 (0.1)
0.0 (0.3)
⫺0.3 (0.5)
⫺7.4 (0.3)
3.3 (0.6)
0.0083
Pd–Pd
Pd–Au
5.67 (0.06)
1.09 (0.08)
2.741 (0.002)
2.755 (0.002)
5.4 (0.2)
4.4 (0.2)
⫺3.3 (0.2)
⫺2.6 (0.1)
0.0081
Pd–Pd
Pd–Au
1.91 (0.13)
4.12 (0.07)
2.744 (0.003)
2.794 (0.002)
4.4 (0.3)
5.5 (0.2)
⫺3.9 (0.2)
⫺7.6 (0.6)
0.0292
Pd–Pd
12.00 (0.12)
2.748 (0.004)
6.4 (0.5)
2.5 (0.3)
0.0043
22 ⫻ 10
Pdnuclei
Pdnuclei–Austack-1
NPs
(Pdnuclei–Austack 1)–
Pdsurf NPs
79.2 ⫻ 10⫺5
Pdnuclei–Austack-2
NPs
(Pdnuclei–Austack-2)–
Pdsurf NPs
6.8
8.6
96.8 ⫻ 10⫺5
Pdnuclei–Austack-3
NPs
0.862
10.4
14.7
(Pdnuclei–Austack-3)–
Pdsurf NPs
110 ⫻ 10⫺5
(Pdnuclei–Austack-1)–
(Pd–Aualloy)surf NPs
110 ⫻ 10⫺5
0.997
16.5
1.000
Pd foil
a
R factor
N, coordination number; Rj, coordination distance; ␴ , Debye–Waller factor; ⌬E0, inner potential correction.
2
FT-EXAFS data analysis disclosed that Pd ions are
completely reduced in all the HRRs, and the Au LIIIedge FT-EXAFS data results revealed that Au ions are
completely reduced during all the RTMs.
The k3-weighted EXAFS oscillations at the Pd K-edge
recorded for (Pdnuclei–Austack-1)–Pdsurf, (Pdnuclei–
Austack-2)–Pdsurf, and (Pdnuclei–Austack-3)–Pdsurf NPs
formed during HRRs, and at the Au LIII-edge recorded
for Pdnuclei–Austack-1, Pdnuclei–Austack-2, and Pdnuclei–
Austack-3 NPs formed during RTMs, are shown in Figures 5 and 6, respectively. The EXAFS oscillation frequencies and the features of the curves at both the Pd
K-edge and the Au LIII-edge of each state are different
from those of Pd and Au foils, especially in the high-k region, indicating the existence of a correlation between
Pd and Au. All these observations indicate that the metallic ions are reduced and the Pd–Au bimetallic clusters are formed in each state.
The reliability of the XAS data obtained at all the
stages in the formation of Pd–Au bimetallic clusters is
checked by comparing the FEFF7 theoretical fit with the
back-transformed experimental EXAFS data obtained
at the Pd K-edge and the Au LIII-edge. However, we
have shown the fitting data only for the bimetallic
Pd–Au clusters with various atomic stackings, i.e.,
(Pdnuclei–Austack-1)–Pdsurf, (Pdnuclei–Austack-2)–Pdsurf,
and (Pdnuclei–Austack-3)–Pdsurf NPs formed at the three
atomic ratios (NAu/NPd) 0.633, 0.862, and 0.997, respectively. The two-shell theoretical fit matches closely with
the back-transformed experimental data (shown as
open circles and solid squares, respectively, in Figures
118
⌬E0 (eV)
VOL. 1 ▪ NO. 2 ▪ CHEN ET AL.
2S and 3S in the Supporting Information), indicating
satisfactory fitting for not only the Pd–Au and Pd–Pd
bonds but also the Au–Pd and Au–Au bonds.
Architecture and Formation of Bimetallic Pd–Au Bimetallic
Nanoparticles in AOT Reverse Microemulsions. In this section we
will discuss the structural details that are obtained
fromthe XAS parameters obtained at the Pd K-edge
and the Au LIII-edge, shown in Tables 1 and 2, respectively. A schematic representation of the architecture of
Pd–Au bimetallic NPs formed at various stages is shown
in Scheme 1.
HRR-1. The coordination number NPd–Cl of the K2PdCl4
emulsion is found to be 4.00 in the absence of reducing
agent. After the addition of 22 ⫻ 10⫺5 mol of N2H5OH
at 2.3 h, NPd–Cl is decreased to 1.86, and contribution
from Pd coordination begins to appear (NPd–Pd, 6.61),
indicating the formation of Pd nuclei through the HRR
enhanced by an intermicellar exchange process.42 The
reduction of metal ions within the cores of reverse
micelles can result in the growth of nano-sized metal
particles.30–33 This particle growth within the cores of
the reverse micelles depends strongly on the exchange
of the reactants between micelles. As can be seen,
further increasing the hydrazine dosage to 44 ⫻ 10⫺5
mol at 4.5 h leads to a complete absence of Pd–Cl
coordination and an increase in Pd–Pd coordination,
indicating the growth of Pdnuclei clusters (see HRR-1,
Scheme 1). The coordination number derived from XAS
is a strong and nonlinear function of the particle
diameter up to 3–5 nm. This property has been widely
used in EXAFS analysis to determine the particle size.26
www.acsnano.org
Pd–Au Bimetallic Clusters with Various Atomic Stackings and the Reference Au Foila
reaction
state
amount of
N2H5OH (mol)
Pdnuclei–Austack-1
NPs
(Pdnuclei–Austack-1)–
Pdsurf NPs
79.2 ⫻ 10⫺5
atomic ratio
(NAu/NPd)
6.0
0.633
7.3
shell
N
Rj (Å)
␴2
(ⴛ10ⴚ3 Å2)
⌬E0 (eV)
R factor
Au–Pd
Au–Au
1.05 (0.12)
7.95 (0.12)
2.755 (0.002)
2.855 (0.003)
11.3 (0.7)
8.7 (0.3)
⫺4.6 (0.8)
2.7 (0.2)
0.0154
Au–Pd
Au–Au
1.37 (0.14)
8.17 (0.14)
2.755 (0.003)
2.855 (0.004)
10.2 (0.8)
9.5 (0.4)
⫺4.6 (0.7)
2.9 (0.5)
0.0196
Pdnuclei–Austack-2
NPs
11.7
0.862
Au–Pd
Au–Au
1.25 (0.11)
8.45 (0.09)
2.755 (0.002)
2.855 (0.003)
14.8 (0.5)
8.2 (0.3)
⫺4.6 (0.7)
2.0 (0.2)
0.0083
Pdnuclei–Austack-3
NPs
17.8
0.997
Au–Pd
Au–Au
1.10 (0.09)
9.20 (0.10)
2.755 (0.001)
2.856 (0.004)
14.7 (0.6)
8.4 (0.2)
⫺4.6 (0.7)
2.1 (0.2)
0.0081
1.000
Au–Pd
Au–Au
4.12 (0.15)
6.41 (0.09)
2.794 (0.003)
2.822 (0.004)
0.5 (0.2)
3.1 (0.3)
7.4 (0.5)
10.8 (0.3)
0.0292
Au–Au
12.00 (0.15)
2.865 (0.005)
8.4 (0.5)
1.1 (0.4)
0.0120
(Pdnuclei–Austack-1)–
(PdAu)alloy NPs
110 ⫻ 10⫺5
Au foil
a
time
(h)
ARTICLE
TABLE 2. Structural Parameters Derived from the Au LIII-Edge EXAFS Data Analysis at Various Stages in the Formation of
Notation as in Table 1.
The Pd–Pd coordination number of Pdnuclei clusters is
found to be 8.9, indicating that the diameter of Pd
clusters is between 1.5 and 2.0 nm.
RTM-1. After the addition of a AuCl4⫺ emulsion to the
microemulsion containing Pd clusters, the NPd–Cl, NPd–Pd,
and NPd–Au are found to be 2.84, 2.17, and 0.21, respectively (NAu/NPd ⫽ 0.633). The decrease in NPd–Pd coordination and increase in NPd–Cl coordination suggests the sacrificial oxidation of Pd clusters upon interaction with
Au3⫹ ions, and the appearance of NPd–Au indicates the
deposition of Au on the Pd surface via a redox transmetalation process to form Pdnuclei–Austack-1 NPs with a first
Au stacking (RTM-1, Scheme 1). As two kinds of chemical species, PdCl42⫺ and Pd–Au clusters, with three kinds
of chemical bonds, Pd–Cl, Pd–Pd, and Pd–Au, dispersed in
the solution a three-shell model are used to fit the k3weighted EXAFS data of the Pd K-edge, the real coordination number is underestimated due to the fact that all
the coordination numbers of the Pd–Cl, Pd–Pd, and
Pd–Au are averaged out, since XAS is an averaging technique. This implies not only that the coordination number
of the Pd–Au bond should be larger than 0.21 but also
that the coordination number of the Pd–Cl bond should
be larger than 2.84 at this stage. The higher coordination
of the Pd–Cl bond indicates that most of the Pd clusters
are dissolved and form PdCl42⫺ species. At this stage, we
strongly believe that the electrons required for the reduction of Au3⫹ ions come mainly from the oxidation of Pd
clusters and not from the oxidation of hydrazine. Most of
the hydrazine is consumed during the Pd cluster formation. The amount of residual hydrazine present
in the system is 3 ⫻ 10⫺5 mol, corresponding to a
molar ratio of [N2H5OH]/[Au3⫹] of 0.07:1, far below
the amount required to reduce Au3⫹ ions.43 Besides,
due to the instability of hydrazine, the amount of residual hydrazine should be much less than 3 ⫻ 10⫺5
mol. Hence, we believe that residual hydrazine, if
present, is not sufficient to reduce Au3⫹ ions. The
Figure 5. k3-weighted EXAFS spectra recorded for Pd–Au bimetallic clusters with various atomic stackings at the Pd
K-edge.
Figure 6. k3-weighted EXAFS spectra recorded for Pd–Au bimetallic clusters with various atomic stackings at the Au LIIIedge.
www.acsnano.org
VOL. 1 ▪ NO. 2 ▪ 114–125 ▪ 2007
119
ARTICLE
Scheme 1. Schematic of the architecture of Pd–Au bimetallic NPs in AOT reverse microemulsions.
standard reduction potential of the reaction AuCl4⫺
⫹ 3e⫺ ¡ Au ⫹ 4Cl⫺ (1.008 V vs SHE) is greater than
that of PdCl42⫺ ⫹ 2e⫺ ¡ Pd ⫹ 4Cl⫺ (0.591 V vs
SHE).44 Hence, the AuCl4⫺ ions in the solutions can
be reduced only via the RTM between Pd0 clusters
and Au3⫹ ions, and the electron required for the reduction is contributed from the oxidation of Pd0 to
Pd2⫹. The direct reduction of Au3⫹ ions to Au0 by
the oxidation of Pd0 to Pd2⫹ without any ligand exchange effect can be further seen in the FT k3weighted EXAFS spectral features at the Au LIIIedge and the Pd K-edge of the Pd–Au bimetallic
NPs which are formed by mixing equal numbers of
moles of Pd clusters and AuCl4⫺ precursor, NAu/NPd
⫽ 1 (see Figure 4S, Supporting Information). Two
peaks corresponding to the Pd–Cl and Pd–M coordination in the Pd K-edge spectra and the Au–M coordination in the Au LIII-edge spectra are observed after the addition of an equal number of moles of
AuCl4⫺ precursor to the Pd clusters (Figure 4S, Supporting Information). In our previous study regarding the formation of bimetallic Ag–Pd NPs, the
PdCl42⫺ ions were directly reduced by monometallic Ag NPs via the RTM, and the electron required for
the reduction was contributed from the direct oxidation of Ag0 to Ag⫹.32 We also studied the formation
mechanism of both monometallic Pt31 and bimetallic Pd–Pt NPs33 in AOT reverse micelles. It was found
that the reduction of PtCl62⫺ is not a direct one,
and it involves three steps. The electrons obtained
from the oxidation of Pd0 to PdCl42⫺ species cannot
be used to produce a Pt layer on the Pd clusters directly,33 indicating that the PtCl62⫺ species are not
appropriate candidates for producing Pdcore–Ptshelltype NPs via the RTM. However, in the present study,
the reduction of AuCl4⫺ species is a direct one, suggesting that AuCl4⫺ is a good candidate for controlling the shell layer and further modifying the elec120
VOL. 1 ▪ NO. 2 ▪ CHEN ET AL.
tronic structure and alloying extent of the bimetallic
NPs.
HRR-2. After the first RTM, the dosage of N2H5OH is increased to 79.2 ⫻ 10⫺5 mol at 6.8 h. At this stage, there
is a complete absence of Pd–Cl coordination, and
NPd–Pd and NPd–Au are found to be increased to 6.37
and 0.87, respectively. Also, NAu–Pd and NAu–Au are
found to be 1.37 and 8.17, respectively. These results
suggest that PdCl42⫺ species are completely reduced
back to Pd0 and deposited on the Pdnuclei–Austack-1 NPs,
thereby producing (Pdnuclei–Austack-1)–Pdsurf NPs
(HRR-2, Scheme 1). One can control the surface composition of Pd and Au on the NPs by properly adjusting
the molar ratio of Au3⫹ ions and Pd2⫹ ions in the solution before performing HRRs.
RTM-2. Later, when AuCl4⫺ is added with an NAu/NPd
stoichiometric ratio of 0.862 at 8.6 h to the solution containing (Pdnuclei–Austack-1)–Pdsurf NPs obtained from
HRR-2, NPd–Pd and NPd–Au are decreased to 2.59 and
0.24, respectively, and NPd–Cl is increased to 2.53. These
results indicate that surface Pd atoms from (Pdnuclei–
Austack-1)–Pdsurf stacked NPs are dissolved again and
transformed to the PdCl42⫺ species. The Au3⫹ ions are
reduced to Au0 to form Pdnuclei–Austack-2 with a second
Au stacking (RTM-2, Scheme 1). The increase in coordination NAu–Au at this stage to 8.45 indicates that the
thickness of Au stacking can be controlled by properly
employing the Au precursor solution.
HRR-3. The PdCl42⫺ species generated in the above
step are further reduced to Pd0 after increasing the dosage of N2H5OH to 96.8 ⫻ 10⫺5 mol at 10.4 h. NPd–Pd
and NPd–Au are increased to 6.46 and 1.06, respectively,
and the absence of Pd–Cl coordination is observed.
These results suggest that the Pd0 clusters are deposited on the surface of Pdnuclei–Austack-2 NPs to produce
(Pdnuclei–Austack-2)–Pdsurf NPs.
RTM-3. When Au precursor solution is again introduced into the microemulsion system containing
(Pdnuclei–Austack-2)–Pdsurf NPs formed in the step HRR-3,
www.acsnano.org
NAu/NPd, of Pd–Au Bimetallic Clusters with Various Atomic Stackings
system
molar ratio
(NAu/NPd)
A (ⴝ 冱NAu-i)
B (ⴝ 冱NPd-i)
Pobserved
Robserved
Prandom
Rrandom
JAu (%)
JPd (%)
0.633
9.54
7.24
0.14
0.12
0.61
0.39
22.9
30.8
0.862
9.70
7.52
0.13
0.14
0.54
0.46
24.1
30.4
0.997
10.30
6.76
0.11
0.16
0.50
0.50
22.0
32.0
(Pdnuclei–Austack-1)–
Pdsurf NPs
(Pdnuclei–Austack-2)–
Pdsurf NPs
(Pdnuclei–Austack-3)–
Pdsurf NPs
with an NAu/NPd stoichiometric ratio of 0.997 at 14.7 h,
NPd–Cl, NPd–Pd, and NPd–Au are found to be 1.96, 1.81,
and 0.21, respectively, and NAu–Au is increased to 9.20,
which corresponds to the formation of Pdnuclei–Austack-3
NPs. This observation is similar to what is observed during the formation of Pdnuclei–Austack-1 and Pdnuclei–
Austack-2 NPs, i.e., the occurrence of a replacement reaction between the shell Pd0 clusters and added Au3⫹
ions, except that the increased NAu–Au value suggests
the increased thickness of Au stacking.
HRR-4. Finally, the dosage of N2H5OH is increased to
110.5 ⫻ 10⫺5 mol at 16.5 h in the microemulsion system containing Pdnuclei–Austack-3 NPs formed in the step
HRR-3. NPd–Au and NPd–Pd are found to be 5.67 and
1.09, respectively. The complete absence of Pd–Cl coordination indicates that all PdCl42⫺ species are completely reduced to form (Pdnuclei–Austack-3)–Pdsurf NPs.
Pd and Au Atomic Distribution and Alloying Extent in
(Pdnuclei–Austack)–Pdsurf Stacked Bimetallic Nanoparticles. The
atomic distribution and alloying extent of Pd and Au
in (Pdnuclei–Austack)–Pdsurf stacked bimetallic NPs are
obtained by employing our previously developed XAS
methodology.20 In the case of (Pdnuclei–Austack-1)–Pdsurf
stacked Pd–Au bimetallic NPs formed during the second HRR, NAu–Au and NAu–Pd are determined as 8.17 and
1.37, respectively, giving the total coordination number of Au and Pd around Au, A (⫽ 冱NAu-i ⫽ NAu–Au ⫹
NAu–Pd), as 9.54. Similarly, NPd–Pd and NPd–Au are determined as 6.37 and 0.87, respectively, providing the total coordination number of Pd and Au around Pd, B
(⫽ 冱NPd-i ⫽ NPd–Pd ⫹ NPd–Au), as 7.24. From these values, the structural parameters Pobserved (⫽ NAu–Pd/
冱NAu-i) and Robserved (⫽ NPd–Au/冱NPd-i) are calculated
as 0.14 and 0.12, respectively. The Pobserved and Robserved
values can be used to describe the atomic distribution
of Au and Pd, respectively, in Pd–Au bimetallic NPs. The
alloying extent of Au (JAu) and Pd (JPd) can be estimated from the Pobserved and Robserved values by employing eqs 2 and 3, respectively.
JAu )
Pobserved
× 100 (%)
Prandom
(2)
JPd )
Robserved
× 100 (%)
Rrandom
(3)
Prandom and Rrandom can be taken as 0.5 if the atomic ratio of Au to Pd is 1:1. However, at HRR-2, the atomic rawww.acsnano.org
tio of Au to Pd is calculated as 0.633:1. From this ratio,
Prandom and Rrandom can be taken as 0.61 and 0.39, respectively. By substituting these values into eqs 2 and 3,
JAu and JPd are calculated as 22.9% and 30.8%, respectively. The higher JPd value indicates a higher degree of
alloying around Pd atoms and less segregation of Pd.
In contrast, the lower JAu value suggests a lower degree
of alloying around Au atoms and more segregation of
Au. The JAu and JPd values, along with other structural parameters, for NPs obtained at different reaction states,
i.e., (Pdnuclei–Austack-1)–Pdsurf, (Pdnuclei–Austack-2)–Pdsurf,
and (Pdnuclei–Austack-3)–Pdsurf NPs, are listed in Table 3.
The similarity in JAu and JPd values obtained for different
reaction states indicates that the Pd–Au bimetallic NPs
produced at these reaction states are similar in structure.
For a homogeneous bimetallic Pcore–Rshell cluster
(NP/NR ⫽ 1), for which the core of the cluster is composed of NP atoms of P and the surface is made of NR atoms of R, the total coordination (NP-P ⫹ NP-R) is equal to
12 for the P atom and less than 12 for the R atom.45 If the
total coordination number (NP-P ⫹ NP-R) is much lower
than 12, most of the P atoms are dispersed on the cluster surface. It is very interesting that the total coordination number of Au, A (⫽ 冱NAu-i ⫽ NAu–Au ⫹ NAu–Pd),
gradually increases with the increase in NAu/NPd, indicating that most of the Au atoms are occupied in the core region and only a few Au atoms are dispersed on the cluster surface. However, the total coordination number of
Pd, B (⫽ 冱NPd-i ⫽ NPd–Pd ⫹ NPd–Au) is decreased to 6.76
for (Pdnuclei–Austack-3)–Pdsurf NPs, indicating that most of
the Pd atoms are dispersed on the cluster surface. This result is consistent with the UV–vis absorption spectral features shown in Figure 7. As can be seen from Figure 7, the
surface plasmon absorption band of the (Pdnuclei–Austack3)–Pdsurf NPs is much lower than that of the pure Au clusters and closer to that of the pure Pd clusters, indicating
that the surface of the Pd–Au bimetallic NPs obtained in
this work has more Pd atoms than the inner core. Both
XAS and UV–vis observations show that the Pd–Au bimetallic NPs follow (Pdnuclei–Austack)–Pdsurf stacking, and
only a few Au atoms are dispersed on the catalyst surface.
A HRTEM image of (Pdnuclei–Austack-3)–Pdsurf NPs is
shown in Figure 8. The image shows that the size distribution of the clusters is monodisperse. The particle
size of clusters obtained from the TEM image, about
2.5–3.0 nm, is slightly larger than that obtained from
XAS analysis (ca. 2.5 nm). This may happen due to the
aggregation of particles caused by the drying process
VOL. 1 ▪ NO. 2 ▪ 114–125 ▪ 2007
ARTICLE
TABLE 3. Structural Coordination Number Parameters and Alloying Extent of Au and Pd for Different Atomic Ratios,
121
ARTICLE
Figure 7. UV–vis spectra of pure Au, Pd, (Pdnuclei–Austack-3)–
Pdsurf, and (Pdnuclei–Austack-1)–(Pd–Aualloy)surf NPs.
before HRTEM analysis. Our previous studies on Pt formation in AOT reverse micelles too indicated that the
particle size obtained from TEM is larger than that obtained from XAS analysis.31
In order to check the applicability of the present design methodology, a bimetallic (Pd–Au)alloy cluster with a
(Pdnuclei–Austack-1)–(Pd–Aualloy)surf stacking was synthesized. At first the Au3⫹ ions with NAu/NPd ⫽ 1 were added
to the preformed Pdnuclei cluster. The RTM between Pdnu3⫹
clei clusters and the Au
ions generated Pdnuclei–Austack-1
2⫺
NPs, leaving PdCl4 and residual AuCl4⫺ species. The
general idea in adding excess Au precursor to the Pdnuclei clusters is to control the insufficient amount of the
electron donated by the surface Pd atoms of the Pdnuclei
clusters required for the reduction of Au precursor. In the
corresponding k3-weighted EXAFS spectra at the Au LIIIedge and at the Pd K-edge, two peaks were observed, corresponding to the coordination of Pd–Cl and Pd–M (in
Pd K-edge EXAFS spectra) and the coordination of Au–Cl
and Au–M (in Au LIII-edge EXAFS spectra) (see Figure 4S,
Supporting Information). Later, in order to make a (Pd–
Au)alloy layer on the surface of Pdnuclei–Austack-1 NPs, HRR
was performed by adding 110 ⫻ 10⫺5 mol of N2H5OH
into the microemulsion system to allow the co-reduction
of both the Pd2⫹ and Au3⫹ ions. At this stage, the Pd–M
bond at the Pd K-edge and the Au–M bond at the Au LIIIedge were split into two peaks, located between 2.0 and
3.0 Å, and both the Pd–Cl and Au–Cl bonds disappeared.
These results indicate that the cluster possesses an atomic
distribution in the shell region different from that of the
(Pdnuclei–Austack)–Pdsurf stacked bimetallic NPs. The fitting
of the (Pdnuclei–Austack-1)–(Pd–Aualloy)surf NPs at the Pd
K-edge and the Au LIII-edge, as shown in Tables 1 and 2,
respectively, indicates that the two peaks at the Pd
K-edge correspond to the Pd–Pd and Pd–Au bonds, and
the two peaks at the Au LIII-edge correspond to the
Au–Pd and Au–Au bonds. Here, too, the Pd–Pd and
Pd–Au bonds at the Pd K-edge, and the Au–Pd and
Au–Au bonds at the Au LIII-edge, are reliable as realized
from the close matching of the two-shell theoretical fit
122
VOL. 1 ▪ NO. 2 ▪ CHEN ET AL.
Figure 8. HRTEM image of the (Pdnuclei–Austack-3)–Pdsurf NPs
in AOT reverse microemulsions.
with that of the back-transformed experimental spectra
(see Figure 5S, Supporting Information). The composition
of the designed (Pdnuclei–Austack-1)–(Pd–Aualloy)surf NPs is
calculated from XAS by measuring the edge jump at both
the Pd K-edge and the Au LIII-edge, and NAu/NPd is found
to be 1.00. NAu–Au and NAu–Pd are found to be 6.41 and
4.12, respectively and the total coordination number, A
(⫽ NAu-i), is 10.53. NPd–Pd and NPd–Au are found to be 1.91
and 4.12, respectively, and the total coordination number, B (⫽ NPd-i), is 6.03. By using eqs 2 and 3, JAu and JPd
are calculated as 78.0% and 136%, respectively. The JAu
and JPd values of (Pdnuclei–Austack-1)–(Pd–Aualloy)surf NPs
show an increasing trend when compared to the JAu and
JPd values of (Pdnuclei–Austack-1 to 3)–Pdsurf NPs. We believe
that the co-reduction of Pd2⫹ and Au3⫹ ions generates a
higher population of alloyed Pd and Au and thereby increases JAu and JPd. That JPd ⬎ 100% and JAu ⬍ 100% indicates that the “Au” atoms prefer “Au” rather than “Pd”
and the “Pd” atoms prefer “Au” rather than “Pd”, and as an
average result, the atomic distribution of the Pd atoms is
better than that of the Au atoms. The UV–vis spectra
shown in Figure 7 show that the surface plasmon absorption band of the (Pdnuclei–Austack-1)–(Pd–Aualloy)surf NPs is
slightly higher than that of the Pd–Au bimetallic clusters,
indicating that the surface is made of a Pd–Au alloy layer.
This observation indicates that (Pdnuclei–Austack-1)–
(PdAualloy)surf NPs can be successfully built by modifying
the surface of Pdnuclei–Austack-1 stacked bimetallic NPs
and further supports the possibility of controlling the surface composition of Pd and Au in the cluster. In the resultant cluster, Au atoms are rich in the core region, alloyed
Pd–Au atoms are preferentially located on the surface
with a higher atomic distribution of Pd, and Pd atoms act
as nuclei to form (Pdnuclei–Austack-1)–(Pd–Aualloy)surf NPs.
The effective control of the RTMs and HRRs can really offer the design and construction of bimetallic clusters with
controlled properties in microemulsions. The present
XAS methodology can be easily extended to control the
alloying extent, atomic distribution, atomic stacking, and
electronic structure of other bimetallic NP systems.
www.acsnano.org
NPs, respectively. We believe that the surface composition of Pd and Au in Pd–Au bimetallic NPs can be controlled by properly employing hydrazine, Pd, and Au
precursor solutions. This is of considerable importance,
since the sizeable (Pd–Aualloy)surf may impart superior
catalytic properties to (Pdnuclei–Austack-1)–(Pd–Aualloy)surf NPs when compared to the Pdnuclei–Austack-1 and
(Pdnuclei–Austack-1)–Pdsurf stacked NPs. The Pd–Au NPs
with variable Au stacking and sizeable surface Pd and
Au populations may be found useful for catalytic applications. Our ongoing work is aimed at modifying the
Pdsurf layer of (Pdnuclei–Austack-1)–Pdsurf stacked bimetallic Pd–Au NPs systematically with the other possible
metals, to generate other bimetallic systems for interesting applications, and investigating their structures
by the described XAS methodology.
EXPERIMENTAL SECTION
(Pdnuclei–Austack-2)–Pdsurf Nanoparticles via Hydrazine Reduction Reaction 3
(HRR-3). Again we re-reduced the Pd2⫹ ions on the surface of
Pdnuclei–Austack-2 NPs formed in step RTM-2 by adding 96.8 ⫻
10⫺5 mol of N2H5OH at 10.4 h to generate (Pdnuclei–Austack-2)–
Pdsurf NPs.
(Pdnuclei–Austack-3) Nanoparticles via Redox Transmetalation Reaction 3 (RTM3). To generate a third stacking of the Au layer, Au3⫹ ions were
added with an NAu/NPd ratio of 0.997 at 14.7 h to the solution obtained in step HRR-3, causing the RTM between the Pdsurf particles and Au3⫹ ions to take place to form Pdnuclei–Austack-3 NPs
with an increase in Au layer thickness.
(Pdnuclei–Austack-3)–Pdsurf Nanoparticles via Hydrazine Reduction Reaction 4
(HRR-4). We further re-reduced the Pd2⫹ ions on the surface of
Pdnuclei–Austack-3 NPs formed at RTM-3 by adding 110 ⫻ 10⫺5
mol of N2H5OH at 16.5 h to generate (Pdnuclei–Austack-3)–Pdsurf
NPs.
The cycle of hydrazine reduction–redox transmetalation reactions can be continued to generate NPs with various atomic
stackings. Meanwhile, the surface composition of NPs can be manipulated by carrying out the HRRs in the microemulsion solution containing a proper ratio of Pd2⫹and Au3⫹ ions which can
be added.
In Situ XAS Measurements. The X-ray absorption spectra at the Pd
K-edge and the Au LIII-edge were recorded at beamline BL12B2 at
the Spring-8, Japan, and at beamline BL01C1 at the National Synchrotron Radiation Research Center, Hsinchu, Taiwan. For beamline
BL12B2, the electron storage ring was operated at an energy of 8
GeV and a beam current of 100 mA. A double-crystal Si(111) monochromater was employed for energy selection, with a resolution
⌬E/E better than 1 ⫻ 10⫺4 at both the Pd K-edge (24352 eV) and
the Au LIII-edge (11917 eV). For beamline BL01C1, the electron storage ring was operated at 1.5 GeV with a beam current of 300 mA.
BL01C1 utilizes the radiation from a superconducting wavelength
shifter (Ec ⫽ 7.5 keV). The beamline employs a double-crystal
Si(111) monochromator for energy selection, with a resolution
⌬E/E better than 2 ⫻ 10⫺4 at both the Pd K-edge and the Au LIIIedge. Higher harmonics were eliminated by detuning doublecrystal Si(111) by using a Rh-coated mirror at beamline BL12B2 or
a Pt-coated mirror at beamline BL01C1. The formation of Pdnuclei
clusters and corresponding Pd–Au bimetallic clusters with various
atomic stackings in reverse micelles was carried out in a homemade cell made of poly(tetrafluoroethylene) for in situ XAS study.
Two holes were made in the cell, one on the top and the other on
one side. After the liquid sample was added, the top hole was
sealed with a Teflon rod to avoid exposing the sample to the atmosphere. A hollow Teflon rod with a Kapton film cap at one end
was inserted into the side hole in the in situ cell. The position of
the Teflon rod was adjusted to reach the optimum absorption
thickness (⌬␮x ⬇ 1.0, where ⌬␮ is the absorption edge and x is
the thickness of the liquid layer) so that the proper edge-jump step
Synthesis of Nanoparticles in AOT Reverse Micelles. Pdnuclei Nanoparticles
via Hydrazine Reduction Reaction 1 (HRR-1). The monometallic Pdnuclei
clusters, used as both templates and seeds in further reactions,
were prepared in the microemulsion solution containing
n-heptane as the continuous oil component, water, and surfactant. The surfactant used was 1 M sodium bis(2ethylhexyl)sulfosuccinate (AOT) (99%). The n-heptane and the
surfactant were thoroughly mixed, and an aqueous solution of
0.82 mL of 0.5 M K2PdCl4 was subsequently added to form a welldefined microemulsion phase with Pd complex in the water
pool. The volume ratio of an aqueous phase to an organic phase
was kept at 1:10. An important parameter characterizing the microemulsion, i.e., the water-to-surfactant molar ratio, W0 {(water)/
(surfactant)}, was equal to 5.5 in the present experiment. This microemulsion solution was placed in the liquid cell, which was
carefully designed governing the edge jump, for the in situ XAS
measurement. A microemulsion of the same composition of oil,
water, and the surfactant that contained the reducing agent was
also prepared. The reducing agent solution was composed of 1
M hydrazine (N2H5OH). To control the particle size of the Pd clusters, an appropriate amount of reducing agent containing microemulsion was then gradually added to the microemulsion containing Pd complex using a microsyringe each hour, starting
from 1 and going to 4.5 h, whereby the Pd complex was reduced to Pd0 clusters.
(Pdnuclei–Austack-1) Nanoparticles via Redox Transmetalation Reaction 1 (RTM1). To the microemulsion containing Pdnuclei clusters was introduced the Au precursor solution, i.e., the HAuCl4 emulsion, which
was prepared by using a procedure similar to that used for the
preparation of the K2PdCl4 microemulsion, at 5.5 h with a NAu/NPd
ratio kept at 0.633 to allow the first RTM between Pdnuclei clusters
and the added Au3⫹ ions. In this reaction, part of the Pd from the
preformed Pdnuclei clusters was oxidized to Pd2⫹, and Au3⫹ ions
were reduced on Pdnuclei clusters, producing Pd–Au bimetallic NPs
with a first atomic stacking (Pdnuclei–Austack-1).
(Pdnuclei–Austack-1)–Pdsurf Nanoparticles via Hydrazine Reduction Reaction 2
(HRR-2). To the solution obtained in step RTM-1, containing
Pdnuclei–Austack-1 NPs and Pd2⫹ ions, the dosage of hydrazine
was increased to 79.2 ⫻ 10⫺5 mol at 6.8 h in order to perform a
HRR. The Pd2⫹ ions in the microemulsion solution were reduced
back to Pd on the surface of the Pdnuclei–Austack-1 NPs, and
(Pdnuclei–Austack-1)–Pdsurf stacked NPs were formed.
(Pdnuclei–Austack-2) Nanoparticles via Redox Transmetalation Reaction 2 (RTM2). Au3⫹ ions with an NAu/NPd ratio of 0.862 were added at 8.6 h
into the solution obtained in step HRR-2, containing the
(Pdnuclei–Austack-1)–Pdsurf NPs. The Pdnuclei–Austack-2 NPs were
produced via the RTM between Pdsurf particles and Au3⫹ ions.
The thickness of the Austack layer of the Pdnuclei–Austack-2 NPs was
increased.
www.acsnano.org
VOL. 1 ▪ NO. 2 ▪ 114–125 ▪ 2007
ARTICLE
CONCLUSION
We have successfully designed and controlled the
nucleation and growth of Pd–Au bimetallic NPs with a
(Pdnuclei–Austack)–Pdsurf stacking within AOT reverse microemulsions as monitored by in situ XAS. The overall
architecture of Pd–Au bimetallic NPs was reasonably explained on the basis of the in situ XAS observations.
We have demonstrated that, by properly employing
the redox transmetalation reaction with a controlled
amount of Au precursor solution, the Au stacking in
Pd–Au bimetallic NPs can be tuned. In addition, we
demonstrated that the surface of Pdnuclei–Austack-1 NPs
can be modified by either Pd or Pd–Aualloy, by employing appropriate hydrazine reduction and redox transmetalation reactions, to generate (Pdnuclei–Austack-1)–
Pdsurf and (Pdnuclei–Austack-1)–(Pd–Aualloy)surf stacked
123
ARTICLE
could be achieved during the measurements. All of the spectra
were recorded at room temperature in transmission mode. Each reaction, from the initial state to the thermodynamically equivalent
state, was monitored endlessly by X-ray absorption spectroscopy.
Pd K-edge XAS spectra were recorded for all the hydrazine reduction reactions (HRR-1 to -4) and redox transmetalation reactions
(RTM-1 to -3), whereas Au LIII-edge XAS spectra were recorded for
all the redox transmetalation reaction steps (RTM-1 to -3) and for
one hydrazine reduction reaction (HRR-2). All reactions were carried out at room temperature. Nitrogen gas was used to purge the
microemulsion system throughout the experiment to prevent the
clusters from being oxidized by air. The incident and transmission
X-ray intensities were respectively detected by ion chambers that
were installed in front of and behind the sample cell. Energy was
scanned from 200 eV below the edge to 1000 eV above the edge.
Standard reference compounds, Pd foil and Au foil for the respective edges, were measured simultaneously by using the third ionization chamber so that energy calibration could be performed
scan by scan.
XAS Data Analysis. Standard procedures were followed to analyze the EXAFS data. First, the raw absorption spectrum in the preedge region was fitted to a straight line, and the background above
the edge was fit using a 11-knot cubic spline function calculated
from AUTOBK code. The EXAFS function, ␹, was obtained by subtracting the post-edge background from the overall absorption
and then normalizing with respect to the edge-jump step. The normalized ␹(E) was transformed from energy space to k-space. The
␹(k) data were weighted by k3 in order to compensate for dampening of the XAFS amplitude with increasing k. Subsequently, k3weighted ␹(k) data in the k-space ranging from 4.05 to 14.10 Å⫺1
for the Pd K-edge and from 3.39 to 13.25 Å⫺1 for the Au LIII-edge
were Fourier transformed to r-space to separate the EXAFS contributions from different coordination shells. A nonlinear leastsquares algorithm was applied to curve-fitting of EXAFS in r-space
between 1.1 and 3.11 Å for the Pd K-edge and between 1.5 and 3.5
Å (without phase correction) for the Au LIII-edge. All the computer
programs were implemented in the UWXAFS 3.0 package,46 with
the backscattering amplitude and the phase shift for the specific
atom pairs being theoretically calculated by using FEFF7 code.47
Palladium and gold foils were employed as references for the
Pd–Pd and Au–Au bonds. Pd–Au and Au–Pd bonds are based on
the face-centered-cubic (fcc) model containing six Pd and six Au atoms in symmetric positions. The amplitude reduction factor, S02,
for Au and Pd was obtained by analyzing the Au and Pd foils, respectively, and by fixing coordination numbers in the FEFFIT input
file. The S02 values are found to be 0.88 and 0.91 for Pd and Au,
respectively.
HRTEM and Spectroscopic Analysis. High-resolution transmission
electron microscopy measurements were carried out using a Phillips/FEI Tecnai 20 G2 S twin transmission electron microscope. The
sample for HRTEM analysis was prepared by placing a drop of the
dispersed solution onto a carbon-coated copper grid and allowing the solvent to be evaporated in a vacuum drybox at room temperature. All the UV absorption spectra were recorded on a Analytik Jena Specord-S 100 UV–vis spectrophotometer.
Acknowledgment. The authors gratefully acknowledge financial support from the National Science Council (NSC-95-2120-M011-002 and NSC-95-2221-E-011-130) and facilities from the National Synchrotron Radiation Research Center (NSRRC), the National
Taiwan University of Science and Technology, and the Institute of
Atomic and Molecular Sciences, Academia Sinica, Taiwan.
Supporting Information Available: k3-weighted EXAFS spectra
of (Pd–Au)alloy clusters at both the Pd K-edge and the Au LIIIedge and their corresponding theoretical fit with backtransformed experimental data. This material is available free of
charge via the Internet at http://pubs.acs.org.
REFERENCES AND NOTES
1. Nagabhushana, K. S.; Bönnemann, H. Wet chemical
synthesis of nanoparticles. In Nanotechnology in Catalysis;
Zhou, B., Hermans, S., Somorjai, G. A., Eds.; Kluwer
Academic/Plenum Publishers: New York, 2004; pp 51–82.
124
VOL. 1 ▪ NO. 2 ▪ CHEN ET AL.
2. Sun, S.; Murray, C. B.; Weller, D.; Folks, L.; Moser, A.
Monodisperse FePt nanoparticles and ferromagnetic FePt
nanocrystal superlattices. Science 2000, 287, 1989–1992.
3. Sinfelt, J. H. Structure of bimetallic clusters. Acc. Chem. Res.
1987, 20, 134–139.
4. Burch, R.; Garla, L. C. Platinum-tin reforming catalyst: II.
Activity and selectivity in hydrocarbon reactions. J. Catal.
1981, 71, 360–372.
5. Shankar, S. S.; Rai, A.; Ahmad, A.; Sastry, M. J. Rapid
synthesis of Au, Ag, and bimetallic Au core–Ag shell
nanoparticles using Neem (Azadirachta indica) leaf broth.
J. Colloid Interface Sci. 2004, 275, 496–502.
6. Damle, C.; Kumar, A.; Sastry, M. Synthesis of Ag/Pd
Nanoparticles and Their Low Temperature Alloying within
Thermally Evaporated Fatty Acid Films. J. Phys. Chem. B
2002, 106, 297–302.
7. Dounda, C. M.; Bertino, M. F.; Blum, F. D.; Tokuhiro, A. T.;
Lahiridey, D.; Chattopadhyay, S.; Terry, J. Radiolytic
Synthesis of Bimetallic Ag-Pt Nanoparticles with a High
Aspect Ratio. J. Phys. Chem. B 2003, 107, 2966–2970.
8. Kumar, D.; Chen, M. S.; Goodman, D. W. Synthesis of vinyl
acetate on Pd-based catalysts. Catal. Today 2007, 123,
77–85.
9. Crathorne, E. A.; MacGowan, D.; Morris, S. R.; Rawlinson,
A. P. Application of isotopic transient kinetics to vinyl
acetate catalysis. J. Catal. 1994, 149, 254–267.
10. Han, Y.-F.; Wang, J.-H.; Kumar, D.; Yan, Z.; Coodman, D. W. A
kinetic study of vinyl acetate synthesis over Pd-based
catalysts: kinetics of vinyl acetate synthesis over Pd–Au/SiO2
and Pd/SiO2 catalysts. J. Catal. 2005, 232, 467–475.
11. Venezia, A. M.; La Parola, V.; Nicoli, V.; Deganello, G. Effect
of gold on the HDS activity of supported palladium
catalysts. J. Catal. 2002, 212, 56–62.
12. Choudhary, T. V.; Sivadinarayana, C.; Datye, A. K.; Kumar,
D.; Goodman, D. W. Acetylene hydrogenation on Aubased catalysts. Catal. Lett. 2003, 86, 1–8.
13. Enache, D. I.; Edwards, J. K.; Landon, P.; Solsona-Espriu, B.;
Carley, A. F.; Herzing, A. A.; Watanabe, M.; Kiely, C. J.;
Knight, D. W.; Hutchings, G. J. Solvent-free oxidation of
primary alcohols to aldehydes using Au–Pd/TiO2 catalysts.
Science 2006, 311, 362–365.
14. Mizukoshi, Y.; Fujimoto, T.; Nagata, Y.; Oshima, R.; Maeda,
Y. Characterization and catalytic activity of core–shell
structured gold/palladium bimetallic nanoparticles
synthesized by the sonochemical method. J. Phys. Chem. B
2000, 104, 6028–6032.
15. Wu, M.; Mao, G.; Meng, M. Preparation and characterization
of the polymer-protected palladium-gold colloidal bimetallic
catalysts. J. Mol. Catal. 1992, 74, 275–284.
16. Schmid, G.; Lehnert, A.; Malm, J.-O.; Bovin, J.-O. Ligandstabilized bimetallic colloids identified by HRTEM and EDX.
Angew. Chem., Int. Ed. Engl. 1991, 30, 874–876.
17. Harpeness, R.; Gedanken, A. Microwave synthesis of core–
shell gold/palladium bimetallic nanoparticles. Langmuir
2004, 20, 3431–3434.
18. Scctt, R. W. J.; Wilson, O. M.; Oh, S. K.; Kenik, E. A.; Crooks,
R. M. Bimetallic palladium– gold dendrimer-encapsulated
catalysts. J. Am. Chem. Soc. 2004, 126, 15583–15591.
19. Kan, C.; Cai, Y.; Li, C.; Zhang, L.; Hofmeister, H. Ultrasonic
synthesis and optical properties of Au/Pd bimetallic
nanoparticles in ethylene glycol. J. Phys. D: Appl. Phys.
2003, 36, 1609–1614.
20. Hwang, B.-J.; Sarma, L. S.; Chen, J.-M.; Chen, C.-H.; Shih, S.C.; Wang, G.-R.; Liu, D.-G.; Lee, J.-F.; Tang, M.-T. Structural
models and atomic distribution of bimetallic nanoparticles
as investigated by X-ray absorption spectroscopy. J. Am.
Chem. Soc. 2005, 127, 11140–11145.
21. Bazin, D.; Guczi, L.; Lynch, J. Anomalous wide angle X-ray
scattering (AWAXS) and heterogeneous catalysts. Appl.
Catal. A: Gen 2002, 226, 87–113.
22. Iwasawa, Y. X-ray absorption spectroscopy for catalysts and
surfaces; World Scientific: Singapore, 1996.
23. Grundwaldt, J. D.; Caravati, M.; Hannemann, S.; Baiker, A.
X-ray absorption spectroscopy under reaction conditions:
suitability of different reaction cells for combined catalyst
www.acsnano.org
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
www.acsnano.org
42. Hirai, T.; Sato, H.; Komasawa, I. Mechanism of formation of
CdS and ZnS ultrafine particles in reverse micelles. Ind.
Eng. Chem. Res. 1994, 33, 3262–3266.
43. Wu, M. L.; Chen, D. H.; Huang, T. C. Synthesis of Au/Pd
bimetallic nanoparticles in reverse micelles. Langmuir
2001, 17, 3877–3883.
44. CRC Hand Book of Chemistry and Physics, 80th ed.; Lide,
D. R., Ed.; CRC Press: Boca Raton, FL, 1999 –2000; p 8.
45. Moonen, J.; Slot, J.; Lefferts, L.; Bazin, D.; Dexpert, H. The
influence of polydispersity and inhomogenity on EXAFS of
bimetallic catalysts. Physica B 1995, 208 –209, 689–690.
46. Stern, E. A.; Newville, M.; Ravel, B.; Yacoby, Y.; Haskel, D.
The UWXAFS analysis package: philosophy and details.
Physica B 1995, 208 –209, 117–120.
47. Zabinsky, S. I.; Rehr, J. J.; Anukodinov, A. L.; Albers, R. C.;
Eller, M. J. Multiple-scattering calculations of x-ray
absorption spectra. Phys. Rev. B 1995, 52, 2995–3009.
VOL. 1 ▪ NO. 2 ▪ 114–125 ▪ 2007
ARTICLE
24.
characterization and time-resolved studies. Phys. Chem.
Chem. Phys. 2004, 6, 3037–3047.
Bazin, D.; Rehr, J. J. Limits and advantages of X-ray
absorption near edge structure for nanometer scale
metallic clusters. J. Phys. Chem. B 2003, 107, 12398–12402.
Bazin, D.; Sayers, D.; Rehr, J. J.; Mottet, C. Numerical simulation of the platinum LIII edge white line relative to nanometer scale clusters. J. Phys. Chem. B 1997, 101,
5332–5336.
Frenkel, A. I.; Hills, C. W.; Nuzzo, R. G. A view from inside:
Complexity in the atomic scale ordering of supported
metal nanoparticles. J. Phys. Chem. B 2001, 105,
12689–12703.
Nitani, H.; Nakagawa, T.; Daimon, H.; Kurobe, Y.; Ono, T.;
Honda, Y.; Koizumi, A.; Seino, S.; Yamamoto, T. Methanol
oxidation catalysis and substructure of PtRu bimetallic
nanoparticles. Appl. Catal. A: Gen. 2007, 326, 194–201.
Sinfelt, J. H.; Via, G. H.; Lytle, F. W. Application of EXAFS in
catalysis. Structure of bimetallic cluster catalysts. Catal.
Rev. Sci. Eng. 1984, 26, 81–140.
Bazin, D. C.; Sayers, D. A.; Rehr, J. J. Comparison between
X-ray absorption spectroscopy, anomalous wide angle Xray scattering, anomalous small angle X-ray scattering,
and diffraction anomalous fine structure techniques
applied to nanometer-scale metallic clusters. J. Phys. Chem.
B 1997, 101, 11040–11050.
Hwang, B.-J.; Tsai, Y.-W.; Sarma, L. S.; Tseng, Y. L.; Liu, D. G.;
Lee, J. F. Genesis of bimetallic Pt–Cu clusters in reverse
micelles investigated by in situ X-ray absorption
spectroscopy. J. Phys. Chem. B 2004, 108, 20427–20434.
Tsai, Y.-W.; Tseng, Y. L.; Sarma, L. S.; Liu, D.-G.; Lee, J.-F.;
Hwang, B.-J. Genesis of Pt clusters in reverse micelles
investigated by in situ X-ray absorption spectroscopy. J.
Phys. Chem. B 2004, 108, 8148–8152.
Chen, C.-H.; Sarma, L. S.; Wang, G.-R.; Chen, J.-M.; Shih, S.C.; Tang, M.-T.; Liu, D.-G.; Lee, J.-F.; Chen, J.-M.; Hwang, B.-J.
Formation of bimetallic Ag–Pd nanoclusters via the
reaction between Ag nanoclusters and Pd2⫹ ions. J. Phys.
Chem. B 2006, 110, 10287–10295.
Chen, C.-H.; Hwang, B.-J.; Wang, G.-R.; Sarma, L. S.; Tang,
M.-T.; Liu, D.-G.; Lee, J.-F. Nucleation and growth
mechanism of Pd/Pt bimetallic clusters in sodium bis(2ethylhexyl)sulfosuccinate (AOT) reverse micelles as studied
by in situ X-ray absorption spectroscopy. J. Phys. Chem. B
2005, 109, 21566–21575.
Wang, D.-Y.; Chen, C.-H.; Yen, H.-C.; Lin, Y.-L.; Huang, P.-Y.;
Hwang, B.-J.; Chen, C.-C. Chemical transformation from
FePt to Fe1-xPtMx (M ⫽ Ru, Ni, Sn) nanocrystal by a cation
redox reaction: X-ray absorption spectroscopic studies.
J. Am. Chem. Soc. 2007, 129, 1538–1540.
Lee, W.-R.; Kim, M. G.; Cho, J.-R.; Park, J.-I.; Ko, S. J.; Oh, S. J.;
Cheon, J. Redox-transmetallation process as a generalized
synthetic strategy for core–shell magnetic nanoparticles.
J. Am. Chem. Soc. 2005, 127, 16090–16097.
Guo, X-C.; Madix, R. J. Oxygen-activated combustion of
alkenes on the Pd (100) surface. J. Am. Chem. Soc. 1995,
117, 5523–5530.
Stuve, E. M.; Madix, R. J. Oxidation of ethylene on Pd (100):
bonding of ethylene and scavenging of dehydrogenation
fragments by surface oxygen. Surf. Sci. 1985, 169, 293–304.
Luo, K.; Yi, C.-W.; Axnanda, S; Goodman, D. W. Preparation
and characterization of silica supported Au–Pd model
catalysts. J. Phys. Chem. B 2005, 109, 23517–23522.
Choi, S. J.; Kang, S. K. Structural transformation of PdPt
nanoparticles probed with X-ray absorption near edge
structure spectroscopy. Catal. Today 2004, 93–95,
561–566.
Prins, R.; Koningsberger, D. C. Catalysis. In X-Ray
Absorption: Principles, Applications, Techniques of EXAFS,
SEXAFS, and XANES; Koningsberger, D. C., Prins, R., Eds.;
Wiley: New York, 1988; p 362.
Materlik, M.; Muller, J. E.; Wilkins, J. W. L-edge absorption
spectra of the rare earths: Assessment of the singleparticle picture. Phys. Rev. Lett. 1983, 50, 267–270.
125